This invention relates to a method of connecting a metallic electrical conductor to an electrically conductive ceramic electrode body to make an electrode assembly which is suitable for use in producing metal by electrolysis.
A number of materials including silicon metals such as aluminum, lead, magnesium, zinc, zirconium, and titanium, for example, can be produced by electrolytic processes. Although individual processes may vary in some respects from one to another, each employs the use of an electrode which must operate in a highly corrosive environment.
An example of such a process for the production of metal is the well-known Hall-Heroult process (hereinafter referred to as the Hall process) for producing aluminum in which alumina dissolved in a molten fluoride salt bath is electrolyzed at temperatures from 900.degree. C. to 1000.degree. C. In the process as generally practiced today, carbon is used as an anode to reduce the alumina, and the reduction produces molten aluminum, and the carbon is oxidized to primarily form CO.sub.2 which is given off as a gas. Despite the common usage of carbon as anode material in practicing the Hall process, there are a number of disadvantages to its use.
Since carbon is consumed in relatively large quantities in the Hall process, approximately 420 to 550 kg per ton of aluminum produced, the anode must be constantly repositioned or replenished to maintain the proper spacing with the cathode in the cell to produce aluminum efficiently. If prebaked anodes are used, it may be seen that a relatively large facility is needed to produce sufficient anodes to operate an aluminum smelter. Furthermore, to produce the purity of aluminum required to satisfy primary aluminum standards, the anode must be relatively pure carbon, and availability and cost of raw materials to make the carbon are of increasing concern to aluminum producers.
Because of the disadvantages inherent in the use of carbon as an anode, there has been a continuing search for inert or nonconsumable materials that can operate as an anode with a reasonable degree of electrochemical efficiency and withstand the high temperature and extremely corrosive environment of the molten salt bath. A number of different types of materials have been suggested and tried, including ceramic oxides, metals and ceramic transition metal borides and carbides, and gaseous fuels, such as natural gas or hydrogen, as the reactant in a fuel-cell type anode. From published literature, few, if any, materials tried will survive for a prolonged time in an aluminum electrolysis cell; however, some ceramic oxides have been reported to be corrosion resistant during cell operation. A recent review of literature and patents relating to inert anodes for use in producing aluminum may be found in articles entitled "Inert anodes for aluminum electrolysis in Hall-Heroult cells (I)" by Kari Billehaug and H. A. Oye, Volume 57, #2, Aluminium, 1981, and "Inert anodes for aluminum electrolysis in Hall-Heroult cells (II)" by Kari Billehaug and H. A. Oye, Volume 57, #3, Aluminium, 1981.
A major problem in the development and use of nonconsumable anodes for producing aluminum by electrolysis has been that of providing a satisfactory method for making a connection between an electrically conductive ceramic material and a metal conductor leading from the cell to a power source. In a typical operation of a Hall cell using carbon as the anode, the anode is formed into a block having a rectangular cross section and a metallic rod or bar is embedded therein by providing a hole in the block, inserting the rod in the hole and filling the void between the rod and the block with molten iron. When the iron solidifies, it shrinks tightly around the bar and away from the hole surfaces of the carbon block, but disengagement is prevented by adapting the block so as to engage the solidified iron. Such an adaptation is providing recesses in the hole side wall, for example. When the above-described assembly is positioned in a Hall cell having a salt bath which is maintained at approximately 1000.degree. C., the rod, cast iron and carbon in the connection zone rise in temperature from room temperature to approximately 700.degree. to 800.degree. C. The rod, cast iron and carbon in the connection zone expand due to this temperature rise and a substantially tight and reasonably efficient electrical connection is effected. Because the rod and cast iron are relatively free to expand longitudinally, the principal electrical contact between the body and the metal due to the thermal expansion is along the lateral surfaces.
When ceramic materials are used for anode bodies, however, such a connection is not satisfactory for a number of reasons.
When using carbon as the anode body, it is desirable that it be in a block form because it is consumed during the electrolytic process and a large block or mass minimizes the frequency with which anodes must be replaced. It is not desirable, on the other hand, to provide an anode of ceramic materials in a large mass or block because, typically, ceramic anode bodies are more expensive to make than are carbon anode bodies, and the carbon materials are typically better conductors of electricity than are ceramic materials used in inert anodes.
As has been previously noted, the carbon anode to metal bar connection utilizing cast iron as the connecting medium relies primarily upon the lateral surfaces of the cast iron being in substantially tight contact with the lateral surfaces adjacent the hole in the carbon block to effect a reasonably satisfactory electrical connection. Variations in electrical conductivity of such a connection due to such things as irregularities in the cast iron and carbon block surfaces, for example, may be tolerated because of the relatively short time span over which an individual carbon block functions as an anode. In the case of an anode made from ceramic materials, however, most of the ceramic materials which are suitable for use as anodes are less efficient electrical conductors than carbon and, furthermore, to be effective, the anode must function over an extended period of cell operation time. Assuring a continuous intimate contact between the ceramic anode body and metal conductor is considered to be more critical, therefore, than the contact required between a carbon block and metal conductor.
Ideally, the connection of a nonconsumable anode material to a metal conductor for use in the electrolytic production of metal must be corrosion resistant, have a minimal voltage drop across the connection, and function to maintain the integrity of the ceramic material when subjected to temperature differentials on the order of 1000.degree. C.
A number of methods for making connections of ceramic materials to metal conductors in the electrolytic production of aluminum have been proposed. Klein U.S. Pat. No. 3,718,550 proposes three different methods. In one of the methods, a ceramic anode tube, having a closed end, contains molten silver and a titanium carbide rod connected to a current supply extends down into the molten silver pool. In a second method, the inner surface of the tube is covered with a thin layer of silver or platinum and a hollow cylinder of nickel-alloy wire mesh is inserted into the tube to contact the silver or platinum layer and is connected with nickel-alloy wires to a conductor leading to the current supply. In the third method, the closed-end ceramic anode tube contains nickel powder, and a rod of zirconium diboride connected to a conductor leading to the current supply is inserted into the nickel powder. Alder U.S. Pat. No. 3,960,678 shows ceramic anode bodies of various shapes in contact with the electrolyte. Adjacent to the anode, but not in contact with the electrolyte, is a material designated as a current distributor which may be a metal such as Ni, Cu, Co, Mo or molten silver or a nonmetallic material such as a carbide, nitride or boride. Power leads connected to the current distributor may be made of the same materials, and it is suggested that the current distributor and power lead may be a single piece. The patentee does not describe how the various connections are to be made. De Nora et al U.S. Pat. No. 4,187,155 suggests attaching lead-in connectors to ceramic electrodes by fusing the connector into the electrode during the molding and sintering process or by making an attachment after sintering, but does not describe any method for making such attachments so as to avoid fracture of the ceramic in use.
Suggestions or descriptions for making metal bonds between ceramics and metals by welding, brazing or other methods of metal bonding have been made. Patents dealing with such methods, for example, are Hackley et al U.S. Pat. No. 3,022,195, Cheng U.S. Pat. No. 3,414,963, Matchen U.S Pat. No. 3,152,871, Zimmer U.S. Pat. No. 3,284,174, Walker U.S Pat. No. 3,839,779, Burgess et al U.S. Pat. No. 3,911,553, Schmidt-Bruecken et al U.S. Pat. No. 3,915,369, Babcock et al U.S. Pat. No. 3,993,411 and Cusano et al U.S. Pat. No. 3,994,430. None of these patents, however, are concerned with connecting an electrically conductive metallic oxide ceramic electrode body to a metal conductor for use in producing a metal by electrolysis. Heretofore, it has not been believed possible to make such a connection in producing a metal because of fracture or failure of the joint due to expansion and/or contraction of the assembly over the extreme temperature differential involved in production of metal by electrolysis.
It would be desirable, therefore, to provide a method for joining a ceramic body to a metal conductor for use in producing metal by electrolysis.